You save $69.98
AZ-400 Premium Bundle
- Premium File 398 Questions & Answers
- Last Update: Nov 11, 2026
- Training Course 27 Lectures
- Study Guide 784 Pages
You save $69.98
A platform team wants to detect repeated failed pipeline stages from a shared template before individual teams escalate.
Which explanation best supports the needed implementation decision?
CI observability complements application monitoring by identifying delivery-system availability and throughput issues. The appropriate response is to collect Azure Pipelines execution status, queue duration and failure trends and alert on actionable thresholds. The expected evidence is that a pipeline failure spike or queue threshold triggers a routed alert with relevant run details.
That explanation is about Kusto query failed requests: Kusto Query Language supports time-window filtering, aggregation and ranking of collected telemetry for focused diagnosis.
This reasoning pertains to Application Insights request telemetry: Application-level instrumentation exposes call outcomes and dependency contributions that basic host availability cannot explain.
That explanation is about Azure Monitor diagnostic collection: Azure Monitor unifies service-specific telemetry needed to distinguish infrastructure faults from application and container behavior.
A research portal team accepts unreviewed direct pushes by contractors even though their work requires approval.
Which explanation best supports the needed implementation decision?
Branch protection enables isolated contribution while enforcing a consistent integration gate before accepted code reaches the default branch. The appropriate response is to use short-lived feature branches and protected pull requests with required reviewers and checks. The expected evidence is that direct pushes are rejected and proposed changes merge only after the configured approvals and successful checks.
This describes the mechanism for Merge policy stale approval: Approvals should cover the final content actually merged, rather than an earlier snapshot of the branch.
This reasoning pertains to Git LFS for binaries: Git LFS replaces eligible large file content in history with lightweight pointers and manages binary content separately.
This describes the mechanism for Trunk-based short branches: Frequent small integrations reduce divergence while flags separate code integration from customer feature exposure.
A release job stores a production database password in plain text in a YAML variable file committed to source control.
Which observation most directly confirms the required DevOps control is effective?
The job accesses the approved secret at runtime and no secret value is present in repository history or published artifacts. This shows that the control was implemented as intended: store sensitive values in Azure Key Vault and retrieve them through authorized pipeline identities at execution time. It matters because a dedicated secret store provides access control, rotation and auditing while preventing credentials from becoming ordinary source artifacts.
That result would verify Managed identity versus service principal, where the relevant control is use a managed identity for the Azure-hosted workload and assign minimal target-resource RBAC roles.
This observation would support Azure Pipelines secure files, where the relevant control is use Azure Pipelines secure files with access restrictions and approved tasks to deliver the file only to authorized jobs.
Such evidence would establish Secret rotation and scope, where the relevant control is separate credentials by environment and purpose, use short lifetimes or federation, and test rotation and revocation.
Two production pipelines deploy different versions of the same service simultaneously and overwrite each other's state.
What is the most appropriate action for this DevOps team?
Configure an exclusive environment lock or equivalent serialized deployment control. Serialization prevents conflicting writes to the same mutable production target while allowing independent builds elsewhere. A useful check is that only one deployment holds the protected environment at a time and later runs follow the agreed queue policy.
This is designed for Pipeline duration hotspot: Critical-path analysis separates true sequential work from jobs that can safely run concurrently.
This would instead address Trigger branch and path filters: Trigger rules map source changes to relevant pipelines and reduce unnecessary executions without weakening critical release checks.
That intervention fits Migrate classic to YAML: YAML makes pipeline implementation reviewable as code, while protected environment checks remain separately configured.
A long-running Azure classic release pipeline has manual settings impossible to review in pull requests.
Which risk most directly follows if the team leaves this problem unresolved?
A rushed rewrite may drop critical approvals or inadvertently change build behavior. The failure follows from the described gap because yAML makes pipeline implementation reviewable as code, while protected environment checks remain separately configured. Mitigate it by choosing to inventory existing classic tasks and resource checks, migrate logic into versioned YAML and validate parity in parallel.
That consequence is characteristic of Build cache with correctness: Correct cache keys improve speed while preventing incompatible or stale cached outputs from silently changing builds. The present scenario instead leaves the migrate classic to yaml requirement unresolved, which creates this more direct risk: a rushed rewrite may drop critical approvals or inadvertently change build behavior.
This risk is associated with Pipeline duration hotspot: Critical-path analysis separates true sequential work from jobs that can safely run concurrently. The present scenario instead leaves the migrate classic to yaml requirement unresolved, which creates this more direct risk: a rushed rewrite may drop critical approvals or inadvertently change build behavior.
That consequence is characteristic of Bicep module reuse: Parameterizing common infrastructure definitions enforces coherent defaults while permitting legitimate environment differences. The present scenario instead leaves the migrate classic to yaml requirement unresolved, which creates this more direct risk: a rushed rewrite may drop critical approvals or inadvertently change build behavior.
A production incident report emphasizes number of completed tickets without showing how long customers lacked service.
Which explanation best supports the needed implementation decision?
Recovery duration captures actual service impact better than activity counts and can guide investment in incident response. The appropriate response is to measure time to restore service from incident impact to verified recovery and trend it alongside incident frequency. The expected evidence is that incident records contain onset and verified recovery times and the restoration trend is shown across comparable events.
This reasoning pertains to Deployment frequency and failure rate: Delivery speed without failure consequences is incomplete; balanced flow and stability indicators reveal unhealthy tradeoffs.
That explanation is about Query-based work metrics: Consistent query definitions produce comparable operational metrics from the underlying tracked work rather than manual snapshots.
This reasoning pertains to Bug quality feedback loop: Tying operational feedback to original code and acceptance criteria exposes recurring failure patterns and supports process improvement.
A GitHub Actions workflow stores a long-lived cloud credential to deploy a production web app even though the platform supports federated login.
What is the most appropriate action for this DevOps team?
Use a scoped workload identity federation service connection or GitHub OIDC trust instead of storing long-lived cloud secrets. OIDC-based federation exchanges trusted workflow identity assertions for short-lived cloud access, reducing static credential exposure. A useful check is that the approved job obtains only scoped Azure access without any stored client secret and untrusted branches cannot impersonate it.
That intervention fits Azure DevOps least privilege: Access levels govern available features while permissions and security groups govern allowed actions; both matter for least privilege.
This is designed for Key Vault runtime secret retrieval: A dedicated secret store provides access control, rotation and auditing while preventing credentials from becoming ordinary source artifacts.
This would instead address GitHub App scope: GitHub Apps separate automation identity from a human account and can grant repository-scoped, revocable access.
An e-commerce app passes functional tests for one user but times out during a seasonal sales surge.
Which risk most directly follows if the team leaves this problem unresolved?
The application appears correct while failing its performance obligations at real traffic volumes. The failure follows from the described gap because load testing exercises concurrency, capacity and bottlenecks that low-traffic correctness tests do not expose. Mitigate it by choosing to run representative load tests with production-like demand and enforce agreed latency and error thresholds.
That is a credible failure of Flaky test quarantine: Flaky tests weaken confidence in build signals and consume delivery capacity; stabilization restores trustworthy pass/fail evidence. The present scenario instead leaves the load test realistic capacity requirement unresolved, which creates this more direct risk: the application appears correct while failing its performance obligations at real traffic volumes.
This risk is associated with Azure App Service deployment slot: Slots provide a separately deployed target and controlled swap that can reduce disruption when application startup is validated. The present scenario instead leaves the load test realistic capacity requirement unresolved, which creates this more direct risk: the application appears correct while failing its performance obligations at real traffic volumes.
That consequence is characteristic of Unit tests fast feedback: Unit tests isolate individual functions or units and provide rapid low-cost correctness feedback before integration environments are needed. The present scenario instead leaves the load test realistic capacity requirement unresolved, which creates this more direct risk: the application appears correct while failing its performance obligations at real traffic volumes.
An engineering manager wants to spot a trend of failures in one shared build agent pool before it halts teams.
Which observation most directly confirms the required DevOps control is effective?
Trend dashboards show rising stage failures and identify the owner and remediation for recurrent issues. This shows that the control was implemented as intended: monitor pipeline success rates, flaky test frequency, queue delay and duration with actionable ownership. It matters because build health includes reliability and time to feedback, not just the final successful outcome.
That result would verify Agent image maintenance, where the relevant control is manage versioned hardened runner images with patch schedules and automated smoke validation.
Such evidence would establish Ring deployment cohort, where the relevant control is implement a ring-based rollout with ordered cohorts, monitoring and explicit promotion criteria.
That result would verify Migrate classic to YAML, where the relevant control is inventory existing classic tasks and resource checks, migrate logic into versioned YAML and validate parity in parallel.
A patch verification job passes locally but fails after a registry adds a newer compatible dependency release.
Which observation most directly confirms the required DevOps control is effective?
Two clean builds of the same commit restore identical package versions and produce equivalent dependency manifests. This shows that the control was implemented as intended: commit a dependency lockfile and use deterministic restore or frozen-lockfile build options. It matters because a lockfile records the resolved dependency graph to make repeated builds use the same versions rather than current registry selection.
This observation would support Semantic versioning, where the relevant control is apply semantic versioning and increase the major version for breaking public compatibility changes.
Such evidence would establish Package feed views, where the relevant control is use Azure Artifacts feed views to promote packages from prerelease into approved release consumption stages.
This observation would support Cross-repository pipeline checkout, where the relevant control is declare authorized repository resources and checkout steps with pinned or reviewed source references.
A service deploys more frequently, but its monitoring dashboard shows no clear link between releases and customer-impacting errors.
Which risk most directly follows if the team leaves this problem unresolved?
Fast delivery can conceal worsening reliability because build success is mistaken for user outcome. The failure follows from the described gap because aligning deployment events to relevant user-facing reliability metrics turns observability into continuous improvement decisions. Mitigate it by choosing to define service-level indicators and use deployment annotations and telemetry-driven post-release feedback.
That consequence is characteristic of Kusto query failed requests: Kusto Query Language supports time-window filtering, aggregation and ranking of collected telemetry for focused diagnosis. The present scenario instead leaves the telemetry slo feedback loop requirement unresolved, which creates this more direct risk: fast delivery can conceal worsening reliability because build success is mistaken for user outcome.
That is a credible failure of Application Insights request telemetry: Application-level instrumentation exposes call outcomes and dependency contributions that basic host availability cannot explain. The present scenario instead leaves the telemetry slo feedback loop requirement unresolved, which creates this more direct risk: fast delivery can conceal worsening reliability because build success is mistaken for user outcome.
This risk is associated with Resource CPU memory investigation: Infrastructure performance indicators help identify resource bottlenecks and distinguish saturation from application defects. The present scenario instead leaves the telemetry slo feedback loop requirement unresolved, which creates this more direct risk: fast delivery can conceal worsening reliability because build success is mistaken for user outcome.
or Guarantee your success by buying the full version which covers the full latest pool of questions. (398 Questions, Last Updated on Nov 11, 2026)
Please check your mailbox for a message from support@examlabs.com and follow the directions.